3245
J . A m . Chem. Soc. 1984, 106, 3245-3252 65-4; (*)-lo (acid chloride), 89578-67-6; (*)-lo (diazo ketone), 89578-81-4; ( f ) - 1 4 , 89578-66-5; (&)-15, 89578-68-7; (f)-15 (acid), 89578-69-8; (*)-16, 89578-70-1; 17,13379-98-1; 17 (triflate), 8957882-5; 18, 89578-71-2; 19, 89578-72-3; 20,89578-73-4; 21,89596-61-2; (*)-23,89578-74-5; (*)-24, 89578-75-6; (&)-24 (2,4,5-trichlorophenol), 89578-83-6; ( f ) - 2 5 , 89578-76-7; (&)-26, 89578-77-8; (*)-27, 8957878-9; (&)-28, 89578-79-0; ( i ) - 2 9 , 89578-80-3; methyl glyoxylate, 922-
68-9; 3-aminopropanol, 156-87-6; N-(4-bromobutyl)phthalimide, 539418-3; trifluoromethanesulfonic anhydride, 358-23-6; 4-aminobutanol, 13325-10-5.
Supplementary Material Available: Complete X-ray d a t a for compound 10 (1 2 pages). Ordering information is given on any c u r r e n t masthead page.
A General Procedure for Preparing a-Lithiosilanes. Generalization of the Peterson Olefination' Theodore Cohen,* James P. Sherbine, James R. Matz, Robert R. Hutchins, Barry M. McHenry, and Paul R. Willey Contribution from the Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260. Received June 23, 1983
Abstract: A particularly convenient method for the preparation of a-lithiosilanes consists of the reductive lithiation of diphenyl thioacetals or thioketals with lithium 1-(dimethylamino)naphthalenide, treatment of the resulting anion with trimethylsilyl chloride, and reductive lithiation, with the same reducing agent, of the resulting a-(pheny1thio)silane. T h e generality of the procedure is demonstrated by t h preparation of a-lithiosilanes in which the negatively charged carbon atom is secondary, tertiary, vinylic, or part of a cyclopropyl ring. These species react with aldehydes and ketones to produce alcohols which, in all cases except the allylic alcohol produced from the vinylic a-lithiosilane, could be induced to form an olefin by loss of the elements of trimethylsilanol upon treatment with potassium hydride or acid.
The Peterson o l e f i n a t i ~ n , involving ~,~~ the reaction of an al i t h i o ~ i l a n with e ~ ~ an aldehyde or ketone followed by elimination of the hydroxide and silicon functions, is a potentially powerful alternative to the Wittig olefination, particularly because of the fact t h a t the elimination step can usually be directed in either a syn or an anti manner.4 However, the method has had the serious limitation that except in special cases a-lithiosilanes have not been readily Since early 19805, we have been studying the feasibility of preparing these organometallics by reductive lithiation, using lithium 1-(dimethylamino)naphthalenide6 (LDMAN), of a-(phenylthio)silanes, a class of compounds two members of which we have reported to be available by reductive lithiation with lithium naphthalenide7 or LDMAN6 of diphenyl thioacetals followed by silylation; in addition t o conventional procedures* for preparing such thioacetals, simple and versatile methods for preparing a variety of cyclopropanone thioketals' and ketene thioacetals1° have recently been developed in our laboratory. Results and Discussion Reparation of a-Lithiosilanes. We now report that this method is completely general for the production of a-lithiosilanes (eq 1 and Table I). We have found LDMAN to be far superior to lithium naphthalenide in the reductive lithiation steps for it obviates the (!) Taken in part (a) from the M.S. thesis of Paul R. Willey, University of Pittsburgh, 1981, and (b) from the Ph.D. thesis of James R. Matz, University of Pittsburgh, 1981. (2) Peterson, D. J. J . Org. Chem. 1968, 33, 780. (3) (a) Chan, T.-d. Acc. Chem. Res. 1977, 10, 442. Colvin, E. "Silicon in Organic Synthesis"; Butterworths: Boston, 1975; Chapters 4 and 12. (b) Paquette, L. A. Science 1982, 21 7, 793. (4) Hudrlik, P. F.; Peterson, D. Tetrahedron Lett. 1974, 1133; J . Am. Chem. SOC.1975, 97, 1464. Hudrlik, P. F.; Peterson, D.;Rona, R. J. J . Org. Chem. 1975, 40, 2263. (5) Many of the results reported here were discussed by T.C. in various lectures during 1981 and late 1980 in the United States and Europe; see footnote 3 of ref 17. (6) Cohen, T.; Matz, J. R. Synth. Commun. 1980, 10, 311. (7) Cohen, T.; Weisenfeld, R. B. J . Org. Chem. 1979, 44, 3601. (8) Grobel, B.-T.; Seebach, D. Synthesis 1977, 357. (9) (a) Cohen, T.;Daniewski, W. M . Tetrahedron Lett. 1978, 2991. (b) Cohen, T.; Weisenfeld, R. B.; Gapinski, R. E. J . Org. Chem. 1979,44,4744. (c) Cohen, T.; Matz, J. R. [bid. 1979, 44, 4816. (10) Cohen, T.;Gapinski, R. E.; Hutchins, R. R. J . Org. Chem. 1979, 44, 3599.
0002-7863/84/1506-3245$01.50/0
I
2
3
4
necessity to separate the naphthalene byproduct from the neutral products; the 1-(dimethy1amino)naphthalene is simply washed out with dilute We have found that even very acid sensitive groups can withstand this treatment.I4 The reductive lithiation method is the only general one for the prepation of a-lithiosilanes containing no additional f~nctiona1ity.l~ An alternative procedure for the production of secondary alithiosilanes 3 (R' # H; R2 = H) involves lithium-selenium exchange, b u t the yields are not entirely satisfactory and t h e method is not applicable to tertiary a-lithiosilanes (3, R' and R2 # H)I6 except for the special case of 1-(1ithiocyclopropyl)trim e t h y l ~ i I a n e . ' ~ J ~The required a-(pheny1thio)silane can be (11) (a) Ager'l has recently reported an analogous sequence for a specific subclass of 1, namely diphenylthioacetals in which R' = H and R2 = phenyl or alkyl, using lithium naphthalenide rather than LDMAN. However, no mention was made of the problem of separating the a-silyl thioether from the naphthalene, one that has caused us and others') great grief before the LDMAN reagent was developed.6 Also in this paperi2 it is implied that treatment of the a-lithiosilanewith ketones or aldehydes yields olefins directly; no details are given. This does not correspond to our experience; as indicated, we found it necessary to treat the alcohols 4 with KH or with acid to effect elimination. (b) Paquette" prepared 3 (R'R2 = CH2CH2)from 2 (R'R2 = CH2CH2)using reductive lithiation with lithium naphthalenide, but the method was abandoned when the chromatography required to separate the alcohols (4, R'R' = CH2CH,) from the naphthalene caused dehydration of the tertiary alcohols. Paquette also claimed that LDMAN gave incomplete reduction; we have found this not to be the case. (12) Ager, D. J. Tetrahedron Lett. 1981, 22, 2983. (13) (a) Paquette, L. A,; Horn, K. A,; Wells, G. J. Tetrahedron Lett. 1982, 23, 259. (b) Paquette, L. A.; Wells, G. J.; Horn, K. A,; Yau, T. H. Tetrahedron 1983, 39, 913. (14) Cohen, T.;Bhupathy, M.; Matz, J. R. J . Am. Chem. SOC.1983,105, 520. ( 1 5) Review of 8-hetero-substituted organometallics: Krief, A. Tetrahedron 1980, 36, 2531. (16) Dumont, W.; Krief, A. Angew. Chem., Int. Ed. Engl. 1976, 15, 161. (17) Halazy, S.; Dumont, W.; Krief, A. Tetrahedron Lett. 1981, 22,4737. (18) A still less general method involves the addition of an alkyllithium to a triorganylvinylsilane. Cason, L. F.; Brooks, H. G. J . Org. Chem. 1954, 19, 1278. Chan, T.-H.; Chang, E.; Vinokur, E. Tetrahedron Lett. 1970, 1137. See also ref 4.
0 1984 American Chemical Society
3246 J . Am. Chem. Soc.. Vol. 106, No. 11, 1984
Cohen et ai.
Table I. Preparation of o-(Pheny1thio)trimethylsilanesand Their Use, via Reductive Lithiation, in the Peterson Olefination w(pheny1thio)trimethylsilane'
7% yieldb
EtCH(SPh)SiMe, (10)
alcohol
% yieldb
86
3od
% olefinbsc
,&-pE'
S Me,
12
11
ri'
MeCh-CtiS8Me3
yield
65g
h
c-C, H,,CH=CMe, 17
84e 62f
13 c-C, HI,CH(SPh)SiMe, (14)
90
i CH
I
Me, (SPh)CSiMe, (15)
92
79
18, R = H 21, R = t-Bu
71 83
51d 5 8 4 m,n
95
85
20, R = H 23, R = t-Bu
9 28'
83,' 59f 800 90p
24 84
86
h
e,.." 86
29
28 92q
9 8s 32
9Oa
lOOf
91"
95f
92
95u'w
35 36 S"e3
%sph
38
40
39
ph?5Me3
OH
85
72x9
41
phxs'Me3 Ph
42 87
2
SPh
48
' Except where otherwise noted, all or(pheny1thio)trimethylsilanes were prepared from the corresponding thioacetal or thioketal. For chromatographed material unless otherwise noted. Yield for elimination step only unless otherwise noted. Considerable decomposition of the alcohol occurred upon flash chromatography. e KH/THF, 19 h; 25 "C. Overall yield of olefin from w(pheny1thio)silane; acidElimination step not performed. ' Reduction to ocatalyzed elimination performed on unpurified alcohol. For distilled product. lithiosilane not performed. Reference 53. Reductive lithiation performed for 4 KH/diglyme, 2.5 h, 90 "C. KH/THF; 0.5 h, 25 "C. min at -78 "C. Overall yield from o-(phenylthio)silane,using KH/THF, 1.5 h, 25 "C, for elimination on unpurified Mixture of epimers. alcohol; yield 89% on the basis of consumed reactant. p KH/THF, 5 h, 9 0 "C. A single isomer of unknown configuration. ' KH/diglyme, 3.5 h , 90 "C. KH/diglyme, 15 min, 25 "C. KH/THF, 1.5 h, 25 "C. Mixture of cis and trans isomers. " It is uncertain if this is one or more geometric isomer. Elimination carried KH/THF, 25 "C. Reaction time: isomer A, 20 h; the mixture of isomers B, C, and D, 4 h. out by Chan's procedure.26d Isomerization product of allene isolated; see text. The substrate did not undergo reductive lithiation in the presence of LDMAN.
'
Preparation of a - Lithiosilanes
J . Am. Chem. SOC.,Vol. 106, No. 11, 1984 3247 vinylcyclopropanone thioketal 7.9bSubsequent formation of the (3-silylcarbinol 39 affords the precursor of 40, a member of an important class of alkylidenecyclopropanes (see below). It would be difficult or impossible to prepare the a-lithiosilane precursor ~ *the ~ ~(3-silylcarbinol ~ (39) by to 39 by competing p r o c e d u r e ~ lor cyclopropanation of a diene.I3
prepared not only from the readily available diphenyl thioacetal group8 as indicated in eq 1 but also by alkylation of the conjugate base of (phenylthio)(trimethylsilyl)methane (2, R’ = RZ = H)]’ and by addition of an alkyllithium to an appropriately substituted alkene. l 9 An attractive feature of the reductive lithiation route to organolithium compounds is the ability to produce tertiary a-lithiosulfides [by reductive lithiation of thioketals 1 (R’ and Rz # H)] and tertiary a-lithiosilanes [3 (R1 and Rz # H)]. Such anionic species are not generally available by deprotonation or other exchange procedures. The instability of these tertiary lithio compounds necessitates certain precautions in their preparation and use and these are indicated in the Experimental Section. A single silylation product was obtained when this procedure was applied to the thioketal of 4-tert-butylcyclohexanone. It is assigned the structure 21 (Table I) on the basis that its precursor anion is protonated exclusively to trans-l-(phenylthio)-4-tertbutylcyclohexane. This result is consistent with the concept that the organolithium is produced by reduction of a radical precursorz0; in other reductions proceeding through tertiary cyclohexyl radicals, the latter apparently assume a conformation in which the substituent on the radical carbon atom assumes the equatorial position.2’ The silylation of the anion derived from the reductive lithiation of 7,7-bis(phenylthio)norcarane affords an epimeric mixture 30 in a ratio of 3:l favoring the endo-7-trimethylsilyl epimer. This assignment is made on the basis that protonation of the same anion affords an identical ratio of exo- and endo-7-(phenylthio)norcarane, both of which are known compounds.zz This can be understood on the basis that there is some preference for the initially produced radicals to interconvert very rapidly and to produce a mixture richer in the less sterically encumbered species possessing an exo substituent.z0b A second electron is donated to form the anion which is configurationally stable.z0b However, on this basis one would expect somewhat similar results in the formation of 35 but a 1:l mixture of epimers is obtained instead. The tertiary a-lithiosilane derived from the reductive lithiation of 21 is protonated or sulfenylated (PhSS0,Ph) to an approximately 1:l mixture of epimers. This may indicate that a flat configuration of the silicon-stablized anion is readily attainable, a conclusion which is consistent with the easily attained linearity of silicon stabilized vinyl anionsz3and the stereochemical instability of 7-lithio-7-(trimethylsilyl)norcaraneat -90 0C.24 Our results show this tendency in the norcarane system as the product of the reaction of the a-lithiosilane derived from the epimers of 30 with an aldehyde afforded a single epimeric alcohol (31 or 33) of unknown configuration. Similarly, the product from the epimers of 35 afforded primarily one epimeric product, 36 (>9:1). These results can be rationalized if it is assumed that the aldehyde approaches the flat silicon-stabilized anions from the accessible exo face of the fused ring systems. Very recently (1 -lithiocyclopropyl)trimethylsilanes have been prepared by selenium-lithiumI7 and bromine-lithium exThe method reported here nicely complements these methods and will be particularly suitable in cases in which the cyclopropanone thioketal is available by one of the newer versatile procedure^.^^^^ An example of such a procedure is the reaction of the sulfur stabilized conjugate base of 5 with 6 to produce the
Only of the previous method^'^-^^ is applicable to the preparation of 9. The ready availability of ketene thioacetals by the reaction of carboxylic acids or esters with aluminum thiophenoxideI0 and by certain Wittig-type connective methods,30as well as the ease of the reductive lithiation step, clearly makes the present procedure of value for preparing silicon-stabilized vinyllithium compounds. It is of some interest to note an apparent minor limitation of this method. The 1-(phenylthio)- 1-(trimethylsilyl)alkene 48, formed in the usual way in two steps from diphenylacetic acid, fails to undergo reductive lithiation with LDMAN; only starting material is recovered. Apparently the radical anion, generated when 48 accepts an electron, has such great stability vis-a-vis the vinyl radical and the thiophenoxide ion that would be produced from it that it fails to irreversibly eject the thiophenoxide ion. Reactions of a-Lithiosilanes with Aldehydes and Ketones. The a-lithiosilanes add in satisfactory yield to aldehydes and to 4tert-butylcyclohexanone (Table I). The a-lithiosilane derived from 24 adds to 4-tert-butylcyclohexanoneat -78 OC to afford an 86% yield of 28. In preliminary experiments using cyclohexanone as the carbonyl compound, the yields of (3-silylcarbinol decreased with increasing reaction temperature. Enolization is the only reaction observed when the a-lithiosilanes derived from 38 and
(19) Ager, D. J.; Cookson, R. C. Tetrahedron Lett. 1980, 21, 1677. Kocienski, P. F. Ibid. 1980, 21, 1559. Ager, D. J. Ibid. 1981, 22, 587. (20) (a) Cohen, T.; Matz, J. R. J . A m . Chem. SOC.1980, 102, 6900. Cohen, T.;Lin, M.-T. Ibid. 1984, 106, 1130. (b) Freeman, P. K.; Hutchinson, L. L. J . Org. Chem. 1980, 45, 3191. (21) House, H. 0.“Modern Synthetic Reactions”; 2nd ed,Benjamin: New York, 1972; pp 155 and 156. (22) Schollkopf, U.; Lehman, G. J.; Paust, J.; Hartl, H.-D. Chem. Ber. 1964, 1527. (23) Zweifel, G.; Murray, R. E.; On,H. P. J . Org. Chem. 1981, 46, 1292. Cunico, R. F. J. Organomet. Chem. 1973, 60, 219. (24) (a) Hiyama, T.; Kanakura, A.; Morizawa, Y.; Nozaki, H. Tetrahedron Lett. 1982, 23, 1279. (b) Brinker, U.H.; Konig, L. J. Am. Chem. SOC. 1979, 101, 4738. (25) We shall soon disclose a new efficient connective method for the preparation of complex 1-(pheny1thio)trimethylsilylcyclopropanes.
(26) (a) Ottolenghi, A.; Fridkin, M.; Silkha, A. Can. J . Chem. 1963, 41, 2977. (b) Grobel, B. T.; Seebach, D. Angew. Chem., Int. Ed. Engl. 1974, 13, 83; Chem. Ber. 1977, 110, 852, 867. (c) Stork, G.; Ganem, B. J . Am. Chem. SOC.1973, 95, 6152. Boeckman, R. K., Jr. Ibid. 1973, 95, 6867. Boeckman, R. K., Jr. Tetrahedron Left. 1974, 3365. (d) Chan, T.H.; Mychajlowskij, W.; Ong.B. S . ; Harpp, D. N. J . Org. Chem. 1978, 43, 1526. (27) Miller, R. B.; McGarvey, G. J . Org. Chem. 1979, 44, 4623. (28) Westmijze, H.; Meijer, J.; Vermeer, P. Tetrahedron Lett. 1977, 1923. (29) Seyferth, D.; Lefferts, J. L.; Lambert, R. L., Jr. J . Organomet. Chem. 1977, 142, 39. These authors reported the preparation of the analogue of 41, in which a bromine atom replaces the PhS group, in 34% yield, by reaction of cyclohexanone with PhHgCBr, and Ph3P followed by n-butyllithium and then Me,SiCI; presumably this compound could be used to generated 9. (30) (a) Seebach, D.; Kolb, M.; Grobel, B.-T. Chem. Ber. 1973, 106, 2277. (b) Mendoza, A.; Matteson, D. S. J . Org. Chem. 1979, 44, 1352.
1
ILLDMAN-CISiMe, r.LDMAN-ArCHO LH,O
39
40
There has been considerable recent interest in the preparation and in a variety of uses of silyl-substituted vinyl anion^.^^,^^ The method of preparation reported here is in most cases a considerable improvement over existing method^^^,^^ which all start from 1silyl- 1-hale1-alkenes, a class of compounds the members of which are sometimes cumbersome to prepare,26a,bx27-z9 Reductive lithiation of I-(phenylthio)-1-(trimethylsily1)alkenes provides an attractive alternative as illustrated by the preparation of 9:
1.LDMAN Wl.10
8
dl
9
3248 J. Am. Chem. SOC.,Vol. 106, No. 11, 1984 21 are treated with the noncyclic ketones acetone and 3-pentanone, respectively, at -78 "C. Elimination to Olefin. The generally preferred method for performing the elimination uses K H in either T H F or diglyme. With KH, eliminations to unstrained olefins and allylidenecyclopropanes often occur at 25 "C in THF. When unconjugated alkylidenecyclopropanes are the products, the eliminations must be performed in diglyme instead of THF. Alcohols 16 and 28, which are particularly resistant to base-induced elimination, give good yields of olefin only when treated with K H in diglyme at 90 "C; it was later found that the base-induced elimination steps generally occur more rapidly in diglyme than in THF. For comparison purposes, certain olefinations were carried out by using H2S04;when 22 was subjected to acid treatment, both the expected olefin 23 and a larger amount of the allylic silane 47 were produced. The one 1-(trimethylsily1)cyclopropylcarbinol
Cohen et al. the synthesis of 40. Alkylidenecyclopropanes are very valuable intermediates in organic synthesis. Oxidation with peracid provides oxaspiropentanes which are capable of rearrangement to cyclobut an one^.^^ Simple alkylidenecyclopropanes have been shown to undergo a variety of metal-catalyzed additions to olefins.3s 2-Vinyl- 1-alkylidenecyclopropanes undergo base-catalyzed rearrangement to 1,3-divinylcyclopropaneswhich, in turn, rearrange to 7-membered rings.36 2-Vinyl- 1-alkylidenecyclopropanes and allylidenecyclopropanes undergo useful thermal rearrangements to 5-member rings.32 Allylidenecyclopropanes undergo DielsAlder reactions with some electrophilic olefins.32i Methylenecyclopropane can be metalated and alkylated directly on the ring.37
Conclusions The reductive lithiation of diphenyl thioacetals and diphenyl thioketals with LDMAN, treatment of the resulting sulfur-stablized anions with trimethylsilyl chloride, and another reductive lithiation is an efficient method for the production of a-lithiosilanes. In this paper for the first time, details are given for the production of both secondary and tertiary a-lithiosilanes by re41 62% 22 23 30% ductive lithiation with this reagent. Precautions that should be observed for the successful preparation of tertiary organolithium species bearing sulfur or silicon on the negatively charged carbon atom are detailed in the experimental section. The procedure can OH . be used for the production of silicon-stablized anions having sp3, cyclopropyl, or sp2 hybridization. The generality of this method, 28 44 the ready availability of starting materials:s the ease of performing the experiments, and the generally satisfactory overall yields 28 which was treated with acid also underwent dehydration to recommend this procedure for a significant role in synthetic ~~~ an allylsilane 44. There are precedents for such b e h a ~ i 0 r . l ~ ~ chemistry. The use of these anions in the Peterson o l e f i n a t i ~ n , ~ ~ However, in certain cases acid-induced eliminations were successful particularly in the preparation of alkylidenecyclopropanes and for the conversion of crude alcohols to the corresponding olefins. allylidenecyclopropanes, is stressed, but there are many other uses This procedure is a significant improvement over the usual olefor such anions; examples are the reactions with dimethylformfination sequence from the a-(pheny1thio)silane for compounds amide to produce a-silylaldehydes which are useful as vinylating whose precursor alcohols are not stable to various chromatographic agents for a n i o r ~ s ,reactions ~ ~ , ~ ~ with sulfur dioxide to produce conditions. Treatment of crude alcohols 11 and 19 with a T H F sulfines?' and in the case of the a-silylvinyl anions, protonation solution of HzSO4 afforded the olefins 12 and 20 in 62% and 59% and alkylation to synthetically useful v i n y l ~ i l a n e s . ~ ~ overall yield from the a-(pheny1thio)silane. Base-induced elimExperimental Section ination was also successfully applied to one of the crude alcohols In the 'H NMR spectra of the (pheny1thio)silanes (2) and p-silyl(see 22, Table I). carbinols (4) the trimethylsilyl group in the molecule was generally used The alcohol 42 posed a problem with regard to elimination. as an internal standard and its chemical shift was defined a t 6 0.00. In Chan26dhas found that analogues of 42 are particularly resistant to the usual methods of elimination, and we could not induce (32) (a) Shields, T. C.; Billups, W. E.; Lepley, A. R. J . Am. Chem. SOC. elimination in the case of 42 by the use of KH. Chanza was able 1968, 90, 4749. (b) Kende, A. S.; Riecke, E. E. Ibid. 1972, 94, 1397. (c) to obtain phenylallene by treatment of the analogue of 42 with Billups, W. E.; Leavell, K. H.; Lewis, E. S.; Vanderpool, S. Ibid. 1973, 95, 8096. (d) Semmelhack, M. F.; DeFranco, R. J. Ibid. 1972, 94, 8838. ( e ) thionyl chloride and subjecting the mixture of allylic chlorides Roth, W. R.; Schmidt, T. Tetrahedron Lett. 1971, 3639. (f) Gilbert, J. C.; to reaction with fluoride ion. All of the examples studied were Higley, D. P. Ibid. 1973, 1075. (g) Billups, W. E.; Baker, B. A.; Chow, W. terminal olefins. When the allylic chlorides obtained by the Y.; Leavell, K. H.; Lewis, E. S. Ibid. 1975, 40, 1702. (h) Huber, M . D.; reaction of 42 with thionyl chloride were treated with tetra-nMartin, R.; Dreiding, A. S. Helu. Chim. Acta 1977,60, 181 1 . (i) Zutterman, F.; Krief, A. J . Org. Chem. 1983, 48, 1 1 35. butylammonium fluoride, a 72% yield of the trans-diene 43 along (33) (a) Bestman, H. J.; Denzel, T. Tetrahedron Lett. 1966, 3591. Sisido, with 15% of a mixture of the two geometric isomers of 46 was K.; Utimoto, K. Ibid. 1966, 3267. Schweizer, E. E.; Berninger, C. J.; realized after chromatography. Diene 46 is the elimination product Thompson, J. G.J . Org. Chem. 1968, 33, 336. Gilbert, J. C.; Weerassoriya, of the chlorides of 42. (46 was the major product obtained by U.; Giamalva, D.Tetrahedron Lett. 1979, 4619. Halazy, S.; Krief, A. Ibid. 1979,4233; 1981,2135. Halazy, S.; Krief, A. J . Chem. Soc., Chem. Commun. treatment of the lithium salt of 42 with p-toluenesulfonyl fluoride.) 1979, 1136. Kitatani, K.; Hiyama, T.; Nozaki, H. J . Am. Chem. Soc. 1975, The expected allene 45 probably undergoes a prototropic shift 97, 949. (b) Hassig, R.; Siegel, H.; Seebach, D. Chem. Eer. 1982, 115, 1990. during the chromatographic procedure to yield 43. A control (34) Crandall, J. K.; Conover, W. W. J . Org. Chem. 1978, 43, 3533 and experiment demonstrated that 43 did not arise by protodesilylation references cited therein. For a recent example of the use of this procedure in synthesis, see: Slessor, K. N.; Oehlschlager, A. C.; Johnson, B. D.; Pierce, of 46.
-
45
42
LQ ,-CH-G-C,H,,
*.To 5 F
43
46
SiMe,
Alkylidenecyclopropanes. The excellent yields of alkylidenecyclopropanes indicate that this method compares favorably with none of which are applicable to extant procedures,17.24.30a~31~32~33 (31) Grobel, B.-T.; Burstinghaus, R.; Seebach, D. Synthesis 1976, 121.
H. D., Jr.; Grewal, S. K.; Wickremesinghe, L. K. G. J. Org. Chem. 1980, 45, 2290. (35) Binger, P.; Cetinkaya, M.; Doyle, M. J.; Germer, A,; Schuchard, U. In 'Fundamental Research in Homogeneous Catalysis"; Tsutsui, M., Ed.; Plenum Press: New York, 1979; p 271. Noyori, R.; Odagi, T.; Takaya, H. J . Am. Chem. SOC.1970, 92, 5780. Albrecht, T. A. J . Organomet. Chem. 1980, 198, 159. Binger, P.; Schuchardt, U. Chem. Ber. 1980, 113, 3334. (36) (a) Billups, W. E.; Shields, T. C.; Chow, W. Y.; Deno, N. C. J . Org. Chem. 1972,37,3676. (b) Billups, W. E.; Chow, W. Y.; Cross, J. H. J . Chem. SOC.,Chem. Commun. 1974, 252. (37) Thomas, E. W. Tetrahedron Lett. 1983, 24, 1467. (38) In a recent paper the inability to prepare 1-lithio-1-(trimethylsily1)ethane for a Peterson olefination step in a total synthesis is lamented: Jung, M. E.; Hudsepth, J. P. J . Am. Chem. SOC.1980, 102, 2463. (39) Unpublished observation of Paul Willey and James Sherbine. (40) Hudrlik, P. F.; Kulkarni, A. K. J . Am. Chem. SOC.1981, 103, 6251. (41) van der Leij, M.; Zwanenburg, B. Tetrahedron Lett. 1978, 3383. (42) Taylor, R. T.; Degenhardt, C. R.; Melega, W. P.; Paquette, L. A. Tetrahedron Lett. 1977, 159.
J . Am. Chem. SOC.,Vol. 106, No. 11, 1984 3249
Preparation of a-Lithiosilanes instances in which this is not the case, it is noted that tetramethylsilane has been added and this was used as the internal standard. Lithium 1-(Dimethy1amino)naphthalenide (LDMAN). To a flamedried 2-neck flask, which was continually purged with argon and was equipped with a glass coated stirring bar, was added T H F (10 mL) and lithium ribbon (40 mg, 5.8 ~ n m o l ) . The ~ ~ mixture was then cooled to between -45 and -55 OC with use of 1-hexanol/dry ice bath. I-(Dimethy1amino)naphthalene (0.84 mL, 0.87 g, 5.1 mmol) was slowly added. The dark green color of the radical anion appeared within 10 min, and the LDMAN was completely formed after 3.5 h of rapid stirring. It is suggested that the LDMAN be used soon after it is formed since at -45 OC it slowly decomposes to 1-lithionaphthalene. If the temperature is maintained below -55 OC, the formation of LDMAN will take considerably more time (Le., -8 h at -78 "C). The above procedure should produce a 0.5 M solution of LDMAN. Generally 2.3 or 2.4 equiv (a 15-20% excess) of LDMAN are used to offset the effect of the lithium oxide formation. Thioacetals and Noncyclopropyl Thioketals. These compounds were synthesized by the procedure of Campaigne and Leal.@ This method involves passing dry HCI gas through a thiophenol solution of the appropriate carbonyl compound or aceta!. 1,l-Bis(phenylthio)cyclohexane was prepared in 87% yield as a white solid: mp 81.5-82.0 OC; IR (KBr) 2930, 1450, 1340, 1100 cm-'; ' H N M R (CCI,) 6 0.9 (s, 9 H, tert-butyl), 1.2-1.9 (m, 9 H, alicyclic), 7.1-7.8 (m, I O H, phenyl). Anal. Calcd for CI8H2$3,: C, 71.95; H, 6.71; S, 21.34. Found: C, 71.65; H , 6.70; S, 21.06. 2,2-Bi~(phenylthio)propane~~ was prepared from the bis(methoxy)propane in quantatative yield as a white solid: mp 56 OC (lit.45mp 56 "C); I R (KBr) 3050,2950, 1480, 1420 cm-I; IH N M R (CCI,) 6 1.0 (s, 6 H, methyl), 7.0-7.3 (m, 10 H , phenyl); mass spectrum (15 eV) 260 (8, M),151 (100,M-SPh),134(12),123(13),117(15),111 (32),73(59). l,l-Bis(phenylthio)propane46 was prepared in 80% yield as an oil from propanal: IR (neat) 3050,2950,2900, 1580, 1475, 1460, 1440 cm-I; 'H N M R (CCI,) 6 1.1 (t, J = 6 Hz, 3 H , methyl), 1.9 (m, 2 H, methylene), 4.3 (t, J = 6 Hz, 1 H , HC(SPh),), 7.0-7.5 (m, 10 H , phenyl); mass spectrum (15 eV) 260 (13, M), 151 (100, M - SPh), 123 (13). l,l-Bis(phenylthio)-4-tert-butylcyclohexanewas prepared in 84% yield as a white solid: mp 82.0-82.5 OC; I R (KBr) 2930, 2850, 1470, 1430, 1360, 1300, 1250, 1100 cm-I; 'H N M R (CCI,) 6 0.9 (s, 9 H, rert-butyl), 1.2-1.9 (m, 9 H, alicyclic), 7.1-7.8 (m, 10 H, phenyl); mass spectrum (15 eV) 247 (100, M - SPh). Bis(phenylthio)cyclohexylmethane was prepared from cyclohexanecarboxaldehyde in quantatative yield as an oil: I R (neat) 2950, 2900, 1590, 1480, 1440 cm-I; 'H N M R (CC,) 6 0.8C-2.20 (m, 11 H, alicyclic), 4.27 (d, J = 2 Hz, 1 H, HC(SPh),), 6.87-7.43 (m, I O H , phenyl); mass spectrum (15 eV) 314 (13, M), 205 (IOO), 123 (38), 95 (69). 1-(Phenylthio)-1-(trimethylsilyl)-4tert -butylcyclohexane (21). A solution of l,l-bis(phenylthio)-4-tert-butylcyclohexane(1.44 g, 4.05 mmol) in 5 mL of T H F was added to a solution of 10.4 mmol of LDMAN in 20 mL of T H F at -78 OC and the resulting mixture was stirred for 15 min. Freshly distilled trimethylsilyl chloride (0.60 mL, 0.51 g, 4.7 mmol) was added to the mixture and within 1 min the following standard workup procedure for non-cyclopropyl reductive lithiations was C ! ' performed. The reaction was quenched with excess water at -78 O The solvent was removed in vacuo, and the residue was taken up in ether. (43) The lithium ribbon becomes coated with oxide while being stored under paraffin oil. Before it is weighed, both sides of the ribbon are scraped free of oxide while the ribbon is still under the oil. The piece of metal is then dipped into pentane to remove the oil and, after rapid air drying, placed in a tared vial of paraffin oil for weighing. However, varying degrees of lithium oxide formation reoccurs. Excessive oxidation can prevent formation of the radical anion. A widely used technique to strip lithium oxide by dipping the lithium into methanol followed by two or three dippings into successive vials of dry THF, then placing it in the reaction vessel, stripped the lithium oxide very well; however, this treatment seemed to deactivate the surface of the metal as radical anion formation was drastically inhibited and the reduction reaction was not complete. It is helpful to tear the lithium ribbon into small pieces in a dry, inert atmosphere prior to placing it in the reaction medium. This provides a fairly fresh area of lithium; scoring the metal is likewise helpful. On occasion, we have experienced difficulty in preparing LDMAN solutions in laboratory conditions of high humidity. (44) Campaigne, E.; Leal, J. R. J . Am. Chem. SOC.1954,76, 1272. (45) Schonberg, A.; Praefcke, K. Chem. Ber. 1967,100, 778. (46) Deljac, A.; Stefanoc, 2.;BahenoviE, K. Tetrahedron Suppl. 1966,8, part I, 33. (47) As discussed above, a slight excess of reducing agent is often used but, when this is the case, it is necessary to quench the reaction immediately after consumption of the anion by silylation since the silylation product can be rapidly reduced by excess radical anion. This "overreduction" is signaled by isolation of some geminal bis(trimethylsily1) compound, and it is only observed with non-cyclopropanone thioketals.
This mixture was washed twice with 5% NaOH, twice with 5% H,SO,, and finally with saturated aqueous NaHCO,. The solvent was removed from the dried (MgSO,) organic layer to yield a crude product. Column chromatography (silica gel, hexanes) afforded 1.08 g (83%) of 21 as a white solid: mp 83.1-83.9 OC; IR (CCI,) 3090, 2950, 1440, 1400, 1370, Me,Si) 6 0.23 (s, 9 H, 1250, 1120, 1020 cm-l; 'H N M R (CCI, CH,Si), 0.80 (s, 9 H, tert-butyl), 0.97-2.00 (m, 9 H , cyclohexyl); high resolution mass spectrum calcd for CI9H3*SSi320.1994, found 320.1994. 1-(Phenylthio)-1-(trimethylsily1)propane(10). l,l-Bis(pheny1thio)propane (10.0g, 38.4 mmol) was treated with LDMAN for 2.5 h, and the resulting a-lithio thioether was silylated and the mixture worked up as usual to afford 7.4 g (86%) of 10 as a clear oil after Kugelrohr distillation (80-85O, 0.25 mm Hg): IR (neat) 3070, 3950, 1580, 1470. 1250 cm-I; 'H N M R (CCI, Me,Si) 6 0.13 (s, 9 H , CH,Si), 1.05 (t, J = 7 Hz, 3 H, CH,C), 1.43-1.97 (m, 2 H , CH2), 2.40 (t, J = 5 Hz, 1 H , HCSPh), 7.03-7.53 (m, 5 H phenyl); high resolution mass spectrum calcd for CI2H2,,SSi 224.1055, found 224.1053. Cyclohexyl(phenylthio)(trimethylsilyl)methane (14). Bis(pheny1thio)cyclohexylmethane (3.93 g, 12.5 mmol) was treated as in the production of 21 to afford 3.15 g (90%) of spectroscopically pure 14 as an oil: IR (neat) 3075, 2925, 2850, 1580, 1470, 1250 cm-'; 'H N M R (CCI, Me,Si) 6 0.17 (s, 9 H, CH3Si), 1.83-2.03 (m, 11 H, cyclohexyl), 2.30 (br d, J = 2 Hz, 1 H, HCSPh), 6.90-7.40 (m, 5 H , phenyl); high resolution mass spectrum calcd for C,,H,,SSi 278.1525, found 278.1523. 2-(Phenylthio)-2-(trimethylsilyl)propane (15). 2,2-Bis(phenylthio)propane (9.94 g, 38.2 mmol) was treated as above to afford 2.96 g (92%) of 15 as an oil after both Kugelrohr distillation (70-80 OC (0.6 mmHg)) and flash chromatography (silica gel, hexanes): IR (neat) 3075, 2960, 1950, 1880, 1460, 1250 cm-l; 'H N M R (CCI, Me,Si) 6 0.13 (s, 9 H, CH3Si), 1.13 (s, 6 H, CMe,), 7.03-7.53 (m, 5 H , phenyl); high resolution mass spectrum calcd for C12H20SSi224.1055, found 224.1055. 1-(Phenylthio)-1-(trimethylsilyl)cyclohexane (18). 1.1 -Bis(phenylthio)cyclohexane (5.70 g, 19.0 mmol) was treated with LDMAN for 1.5 h at -78 "C; it was only after this experiment that it was found that far shorter reaction times provide better yields. (See preparation of 21.) After silylation, the usual workup afforded 1.71 g (71%) of 18 as a white solid, with a room temperature melting point, after flash chromatography (silica gel, hexanes): IR (neat) 3075, 2925, 1580, 1470, 1440, 1260 cm-'; 'H N M R (CCI, Me,Si) 6 0.10 (s, 9 H, CH,Si), 0.73-2.30 (m, 10 H, alicyclic), 6.97-7.53 (m, 5 H, phenyl); high resolution mass spectrum calcd for CI5H2,SSi 264.1 368, found 264.1368. 1-(Phenylthio)-1-(trimethylsily1)cyclopropane (24). A solution of 14.5 g (95.0 mmol) of l,l-bis(phenylthio)cyclopropane9awas treated with LDMAN for 45 min at -45 OC. This is the standard reaction time and temperature for cyclopropanone thioketal reductive lithiations. Trimethylsilyl chloride (19.3 mL, 16.5 g, 15.2 mmol) was added, and the resulting mixture was stirred for 30 ~ n i n . ~ Standard * workup and purification by column chromatography (silica gel, 5% ethyl acetate-hexanes) afforded 20.0 g (95%) of 24 as an oil: IR (neat) 2990, 1595, 1375, 1250, 1090 cm-'; 'H N M R (CDCI,) 6 0.0 (s, 9 H , CH,Si), 1.0 (s, 4 H , cyclopropyl), 7.2-7.6 (m, 5 H , phenyl); high resolution mass spectrum calcd for C,,H,,SSi 222.0899, found 222.0893. This particular a-phenylthiosilane has also been prepared in our laboratory from 1,3-bis(pheny1thio)propane by a method independently used by Paquette.') 6-(Phenylthio)-6-(trimethylsilyl)bicyclo[3.1.O]hexane (35). 6,6-Bis(phenylthio)bicycl0[3.1.0]hexane~~ (2.7 g, 9.2 mmol) was treated with LDMAN and the anion silylated to afford 2.3 g (95%) of 35 after purification by column chromatography (silica gel, hexanes): IH N M R (CDCI,) 6 0.0 ( s , 9 H, CH3Si), 1.4-1.8 (m, 8 H. alicyclic), 6.9-7.1 (m, 5 H, phenyl); high resolution mass spectrum calcd for C,,H2,SSi 262.1212, found 262.1212. 7-(Phenylthio)-7-(trimethylsilyl)bicyclo[4.1.O]heptane (30). 7,7-Bis(phenylthio)bicyclo[4.1.0]heptane50 (9.00 g, 28.8 mmol) was treated in the same way to afford 6.8 g (86%) of 30 after Kugelrohr distillation (145 OC (2 mmHg)): IR (neat) 2950, 1590, 1490, 1440, 1250 cm-I; ' H NMR (CDCI,) 6 0.0-0.3 (s, 9 H, isomeric CH3Si), 1.0-1.8 (m, 10 H, alicyclic), 7.3 (m, 5 H, phenyl); high resolution mass spectrum calcd for C,,H,,SSi 276.1368, found 276.1358. Likewise, 7-(phenylthio)norcarane was formed as an epimeric mixture of sulfides by protonation of the anion prepared as above. The sulfides were separated by medium-pressure chromatography and identified by the coupling constants of the cyclopropyl hydrogen atoms.22
+
+
+
+
+
(48) Bis silylation is not a problem with cyclopropanone thioketa!s. (49) Cohen, T.; Ritter, R. H.; Ouellette, D. J . Am. Chem. SOC.1982,104, 7142. (50) Braun, M.; Seebach, D. Chem. Ber. 1976,109,699.We have found49 that the ring-closure step is best performed with methyllithium rather than butyllithium since the latter causes a significant amount of sulfur-lithium exchange in the product.
3250 J . Am. Chem. Soc., Vol. 106, No. 11, 1984
Cohen et al.
+
exo-7-(Phenylthio)norcarane: 'H N M R (CDCI, Me&) 6 1.208-1.288 (m, 6 H, alicyclic), 1.846-1.876 (m. 2 H, alicyclic), 1.876-1.903 (t. J = 4.05 Hz, 1 H, CH(SPh)), 1.925-2.060 (m, 2 H, alicyclic), 7.090-7.140 (m, 1 H, aromatic), 7.261-7.289 (m, 4 H, aromatic). endo-7-(Phenylthio)norcarane: IH N M R (CDCI, Me,Si) 6 1.255-1.546 (m, 8 H, alicyclic), 1.820-1.960 (m, 2 H, alicyclic), 2.251-2.304 (t, J = 8.0 Hz, 1 H, CH(SPh)), 7.060-7.127 (m,1 H, aromatic), 7.237-7.289 (m, 2 H , aromatic), 7.329-7.361 (m, 2 H, aromatic). cis- and trans -2-(2-Methyl-1-propenyl)-I-( phenylthio)-1-(trimethylsi1yl)cyclopropane(38). 79b(0.35 g, 1.12 mmol) was treated in a similar fashion to afford 0.30 g (95%) of 38 in an epimeric ratio of 2:l after column chromatography: IR (neat) 2950, 2850, 1590, 1490, 1260, 860, 840 cm-l; IH N M R (CDCI,) 6 0.0 and 0.1 (s, 9 H, epimeric CH3Si), 0.8-1.6 (m, 3 H, cyclopropyl), 1.7-1.9 (m, 6 H, isomeric allylic methyl), 4.9-5.4 (m, 1 H, vinyl), 7.0-7.4 (m, 5 H, phenyl); high resolution mass spectrum calcd for C16H2,SSi 276.1 368, found 276.1363. [Cyclohexylidene(phenylthio)methyl]trimethylsilane (41). 71° (1.4 g, 4.5 mmol) was treated with LDMAN for 30 min at -45 OC. To this mixture was added trimethylsilyl chloride (0.68 mL, 0.58 g, 5.4 mmol), and the resulting solution was stirred for 1 h; the long reaction time was used because the vinyl lithium is silylated more slowly than the other anions in this study. This afforded 1.2 g (96%) of spectroscopically pure 41 as an oil: IR (neat) 3075, 2930, 1585, 1570, 1475 cm-'; 'H NMR (CDCI,) 6 0.00 (s, 9 H, CH3Si), 1.34-1.70 (m, 6 H, nonallylic cyclohexyl), 2.23-2.73 (m, 4 H, allylic cyclohexyl), 6.85-7.10 (m, 5 H, phenyl); high resolution mass spectrum calcd for CI6H2,SSi276.1367, found 276.1360. 1-[l-(Trimethylsilyl)propyl]-4-tert-butylcyclohexanol (11). 10 (0.40 g, 1.8 mmol) was treated with LDMAN for 15 min at -78 0C.51 A T H F solution of 4-tert-butylcyclohexanone(0.27 g, 1.78 mmol) was added. After 1 min the mixture was worked up in the usual way. Flash chromatography (silica gel, 2% ethyl acetate-hexane) afforded 0.14 g (30%) of 11 as a pale yellow oil: IR (neat) 3630, 3500, 2950, 1480, 1370, 1255 cm-I; 'H NMR (CCI, Me,Si) 6 0.07 (s, 9 H , CH,Si), 0.53 (t, J = 6 Hz, 1 H, HCSiMe,), 0.83 (s, 9 H, tert-butyl), 0.67-1.87 (m,24 H, alkyl and hydroxyl); mass spectrum (15 eV) 252 (44, M - H 2 0 ) , 180 (63, M - Me,SiOH), 178 (100). 3-(Trimethylsilyl)pentan-2-ol (13). 10 (0.39 g, 1.73 mmol) was treated with LDMAN for 15 min at -78 OC. Acetaldehyde (0.13 mL, 0.10 g, 2.3 mmol) was added (see experiment above). Kugelrohr distillation (75-80 OC, aspirator) afforded 0.18 g (65%) of 13 as a clear oil: IR (neat) 3375, 2950, 1460, 1380, 1110, 1080 cm-l; 'H NMR (CCI,) 6 0.00 (s, 9 H, CH3Si), 0.37-1.77 (m, 9 H, alkyl), 2.00-2.77 (br s, 1 H, OH), 3.67-4.20 (m, 1 H, HCOH); mass spectrum (15 eV) 142 (25, M - H 2 0 ) , 70 (100, M - HOSiMe,). 1-Cyclohexyl-2-methyl-2-(trimethylsilyl)propan-l-ol (16). 15 (0.20 g, 0.91 mmol) was treated with LDMAN for 15 min at -78 OC. Cyclohexanecarboxaldehyde (0.1 1 mL, 0.10 g, 0.91 mmol) was added, and the solution was treated in the usual way. Column chromatography (Silicar-CC-7, 4% ethyl acetate-hexanes) afforded 0.16 g (79%) of 16 as an oil which crystallized on standing: IR (CCI,) 3650, 2950, 1460, 1250, 1080 cm-l; 'H N M R (CCI4 Me,%) 6 0.10 (s, 9 H, CH,Si), 0.93 and 0.97 (two singlets, 6 H, enantiotopic methyls), 0.67-2.00 (m, 12 H , cyclohexyl and hydroxyl), 3.13-3.37 (m,1 H, HCOH). [I-(Trimethylsilyl)cyclohexyl]cyclohexylmethanol (19). 18 (0.47 g, 1.79 mmol) was treated with LDMAN for 15 min at -78 OC. Cyclohexanecarboxaldehyde (0.24 mL, 0.21 g, 1.96 mmol) was added, and the solution was treated in the usual way. Column chromatography (Silicar-CC-7, 1% ethyl acetate-hexanes) afforded 0.24 g (51%) of 19 as an Me,Si) 6 0.03 (s, 9 H, CH,Si), 0.63-2.10 (m, oil: 'H NMR (CCI, 22 H, alkyl and hydroxyl), 3.60 (d, J = 6 Hz, 1 H, HCOH); mass spectrum 250 (3, M - H,O), 178 (100 M - Me,SiOH). (4-Methoxyphenyl)[ 1-(trimetbylsilyl)cyclopropyl]methanol (25). 24 (0.55 g, 2.5 mmol) was treated with LDMAN for 1.5 h at -45 OC. p-Anisaldehyde (0.36 mL, 0.40 g, 3.0 mmol) was added, and the solution was treated in the usual way. Column chromatography (silica gel, 10% ethyl acetate-hexanes) afforded 0.53 g (85%) of 25 as an oil: IR (neat) 3450, 1240, 1040 cm-'; ' H NMR (CDCl,) 6 0.0 (s, 9 H, CH,Si), 0.6-0.7 (m, 4 H, cyclopropyl), 2.2 (br s, 1 H, OH), 4.0 (s, 3 H, OCH,), 4.6 (m, 1 H, CHOH), 6.9-7.5 (m, 4 H, aromatic); high resolution mass spectrum calcd for CI,H2,O2Si 250.1389, found 250.1389. [ 1-(Trimethylsilyl)cyclopropyl]-n- hexylmethanol (27). 24 (0.55 g, 2.5 mmol) was treated with LDMAN for 1.5 h at -45 OC. Heptanal (0.42 mL, 0.34 g, 3.0 mmol) was added, and the solution was treated in the
+
+
+
+
(51) It was later found that a reaction time of 4 min instead of 15 min improved the yield of alcohol for the preparation of 22. See preparation of 23.
usual way. Purification by column chromatography afforded 0.48 g (84%) of 27 as an oil: IR (neat) 3375, 2950, 2900, 1425, 1230, 1100 cm-l., I H N M R (CDCI,) 6 0.0 (s, 9 H, CH,Si), 0.2-0.4 (m, 4 H, cyclopropyl), 1.0-1.8 (m,14 H, hexyl and OH), 3.0 (t, J = 6 Hz, 1 H, HCOH); high resolution mass spectrum calcd for C13H280Si 210.1 804 (M - H 2 0 ) , found 210.1804.
1-[1-(Tnmethylsilyl)cyclopropyl]-4-tert-hutylcyclohexan-l-ol (28). 24 (0.47 g, 2.1 mmol) was treated with LDMAN for 1.5 h at -45 OC. The solution was cooled to -78 OC. A T H F solution of 4-tert-butylcyclohexanone (0.38 g, 2.4 mmol) was added, and the solution was treated in the usual way. Column chromatography afforded 0.43 g (86%) of 28 as a solid (mp 98-100 "C): IR (KBr) 3600, 3500,3000, 1340, 1220, 800 cm-'; 'H NMR (CDCI,) 6 0.0 (s, 9 H, CH,Si), 0.3-0.5 (m, 4 H, cyclopropyl), 0.8 (s, 9 H , tert-butyl), 1.2-1.6 (m, 9 H, alicyclic); high resolution mass spectrum calcd for C,,H,,OSi 268.2222, found 268.2222.
n-Pentyl[7-(trimethylsilyl)bicyclo[4.l.O]hept-7-yl]methanol(31). 30 (0.60 g, 2.2 mmol) was treated with LDMAN for 1.5 h at -45 "C. Hexanal (0.28 mL, 0.23 g, 2.4 mmol) was added, and the solution was treated in the usual way. Column chromatography afforded 0.53 g (92%) of 31 as an oil: IR (neat) 3500, 2950, 2900, 1500, 1280, 860 cm-'; IH NMR (CDCI,) 6 0.0 (s, 9 H, CH,Si), 0.8-1.5 (m, 22 H, alkyl, alicyclic and OH), 3.4 (m, 1 H, HCOH); high resolution mass spectrum calcd for C16H320Si253.1988 (M - CH,), found 253.1988.
2-Propenyl[7-(trimethylsilyl)bicyclo[4.1.0]hept-7-yl]methanoI (33).30 (0.94 g, 3.4 mmol) was treated with LDMAN for 1.5 h at -45 "C. Methacrolein (0.46 mL, 0.38 g, 5.5 mmol) was added, and the solution was treated in the usual way. Column chromatography afforded 0.73 g (90%) of 33 as an oil: IR (neat) 3450, 2950, 1450, 1240, 840 cm-I; ' H NMR (CDCI,) 6 0.0 (s, 9 H, CH,Si), 1.5-2.1 (m, 14 H, alicyclic H, and allylic methyl and OH), 3.8 (m, 1 H, CHOH), 4.9 (m, 1 H, vinyl), 5.2 (m, 1 H, vinyl); high resolution mass spectrum calcd for Cl,H160Si 238.1756, found 238.1746.
2-Propenyl[6-(trimethylsilyl)bicyclo[3.1.O]hex-6-yl]methanol (36). 35 (0.65 g, 2.5 mmol) was treated with LDMAN for 2 h at -45 OC. Methacrolein (0.30 mL, 0.25 g, 3.0 mmol) was added, and the solution was treated in the usual way. Column chromatography afforded 0.51 g (91%) of 36 as an oil: IR (neat) 3450, 2950,2889, 1250, 1120 cm-I; ' H NMR (CDCI,) 6 0.0 (s, 9 H, CH,Si), 1.4-2.2 (m, 12 H, alicyclic H, allylic methyl and OH), 3.8 (m, 1 H, CHOH), 5.0 (m, 1 H, vinyl), 5.3 (m, 1 H, vinyl); high resolution mass spectrum calcd for Cl3H2,0Si 224.1 596, found 224.1596.
cis- and trans-(4-Methoxyphenyl)[2-(2-methyl-l-propenyl)(l-trimethylsilyl)cyclopropyl]methanol (39). 38 (0.10 g, 0.36 mmol) was treated with LDMAN for 40 min at -45 OC. p-Anisaldehyde (0.053 mL, 0.059 g, 0.043 mmol) was added, and the solution was treated in the usual way. Column chromatography (silica gel, 5% ethyl acetate-hexanes) afforded 0.10 g (92%) of 39 as three fractions comprised of four isomers whose RJvalues in 5% ethyl acetate-hexanes are as follows: A (0.29), B (0.17), C (0.12), and D (0.09). These fractions consisted of pure A and pure B along with a mixture of B, C, and D. (It is necessary to note that the products of large-scale reactions (> 1.5 g) should be flash chromatographed as decomposition occurs upon the longer chromatographic times necessary for open column silica gel chromatography; the yield for open column silica gel chromatography is -70%) The ' H NMR spectrum of the mixture is very similar to that of pure A and pure B. The IR and mass spectrum are of a mixture of all of the isomers: IR (neat) 3450, 2950, 2900, 1610, 1510, 1250, 1175, 1040, 840 cm-l. A: ' H NMR (CDCI,) 6 0.00 (s, 9 H, CH,Si), 1.25-1.30 (m, 1 H, cyclopropyl), 1.35-1.40 (m.1 H, cyclopropyl), 1.85-1.95 (m,1 H, tertiary cyclopropyl), 2.05 (5, 3 H, allylic methyl), 2.10 (s, 3 H , allylic methyl), 2.4 (br s, 1 H, OH), 4.05 (s, 3 H, OCH,), 4.75 (m, 1 H, benzylic CHOH), 5.55-5.60 (m,1 H, vinyl), 7.10-7.15 (d, J = 8 Hz, 2 H, aromatic), 7.65-7.70 (d, J = 8 Hz, 2 H , aromatic). B: 'H 6 0.00 (s, 9 H, CH,Si), 0.50-0.60 (m, 2 H, cyclopropyl), 1.50-1.60 (m, 1 H, tertiary cyclopropyl), 1.65 (s, 3 H, allylic methyl), 1.75 (s, 3 H, allylic methyl), 1.80-1.85 (m,1 H, OH), 3.80 (s, 3 H, OCH,), 4.75 (m, 1 H, CHOH), 4.85-4.95 (m,1 H, vinyl), 6.80-6.85 (d, J = 8 Hz, 2 H, aromatic), 7.20-7.25 (d, J = 8 Hz, 2 H, aromatic); high resolution mass spectrum calcd for CI,H2,02Si 304.1859, found 304.1852. l-Cyclohexyl-2-cyclohexylidene-2-( trimethy1silyl)ethan-1-01 (42). 41 (1.9 g, 6.9 mmol) was treated with LDMAN for 1 h at -45 " C . Cyclohexanecarboxaldehyde (0.94 mL, 0.85 g, 7.6 mmol) was added, and the solution was treated in the usual way. Column chromatography afforded 1.6 g (85%) of 42 as an oil: IR (neat) 3550, 2900, 1470, 1270, 1100 cm-l; ' H N M R (CDCI,) 6 0.0 (s, 9 H, CH,SI), 1.4-1.6 (m, 17 H, non-allylic cyclohexyl), 2.0-2.4 (m, 4 H, allylic cyclohexyl), 4.0 (d, J = 8 Hz, 1 H, HCOH); high resolution mass spectrum calcd for C,,HS2OSi 280.2222, found 280.2219. (4-Methoxyphenyl)cyclopropylidenemethane (26). The following standard procedure was used for KH olefinations. In a 2-neck flask,
J . Am. Chem. SOC.,Vol. 106, No. 11, I984
Preparation of a- Lithiosilanes continually being flushed with argon, a solution of 25 (0.090 g, 0.36 mmol) in 10 mL of T H F was stirred as an excess of hexane-washed K H was slowly added through the argon outlet stream (until no further gas evolution was noticed). The reaction mixture was stirred for 20 h at room temperature. The resultant mixture was carefully poured onto ice water overlaid with ether. A preparative TLC plate of the residue of the dried (MgS0,) organic layer, using hexanes as the eluent, afforded 0.060 g (90%) of 26 as an oil whose 'H N M R spectrum was identical with that of an authentic sample:52 'H N M R (CDCI,) 6 1.O-1.5 (m, 4 H, cyclopropyl), 4.0 (s, 3 H, OCH,), 6.6-6.7 (m, 1 H, vinyl), 6.7 (d, J = 8 Hz, 2 H , aromatic), 7.9 (d, J = 8 Hz, 2 H, aromatic). 4-tert-Butyl(propylidene)cyclohexane (12). A solution of 11 (0.10 g, 0.38 mmol) in 5 mL of T H F was treated with K H by the usual procedure. After 19 h the starting material was consumed. Workup by the standard procedure followed by column chromatography (silica gel, hexanes) afforded 0.057 g (84%) of 12 as an oil: IR (neat) 2975, 2875, 2850, 1490, 1425, 1375 cm-I; ' H N M R (CCI,) 6 0.83 (s, 9 H, tert-butyl), 0.63-2.77 (m, 14 H, alicyclic and alkyl), 5.02 (t broadened by allylic coupling, J = 7 Hz, 1 H , vinyl); mass spectrum (15 eV) 180 (100, M), 165'(2< M - CH3). Cvclohexvlidenecvclohexvlmethane (20). A solution of 19 (0.19 g, 0.73"mmol) in 5 mL"of T H F was treated with KH in the usual manner. After 0.5 h at room temperature the starting material was consumed. The reaction was worked up by the usual procedure. Kugelrohr distillation (85-95 "C,aspirator) afforded 0.10 g (83%) of 20 as an oil: IR (neat) 2925, 2850, 1440 cm-I; ' H N M R (CDCI,) 6 0.57-2.57 (m, 21 H, alicyclic), 4.87 (d, broadened by allylic coupling, J = 9 Hz, 1 H , vinyl); mass spectrum (15 eV) 178 (100, M). (4-tert-Butylcyclohexylidene)cyclohexylmethane(23). A solution of 21 (0.20 g, 0.64 mmol) in 1.0 mL of T H F was added to a solution of 1.5 mmol of LDMAN in 3.0 mL of THF, and the resulting mixture was stirred for 4 min at -78 "C. Cyclohexanecarboxaldehyde (0.10 mL, 0.09 g, 0.08 mmol) was added. The resulting solution was stirred for 15 min and worked up by the standard procedure to provide crude 22 which was used directly in the elimination step. In another experiment, a sample of 22 was purified by flash chromatography: I R (CCI,) 3625, 2925, Me,%) 6 0.13 (s, 9 H , 1440, 1335, 1225 cm-'; 'H N M R (CDCI, CH,Si), 0.83 (s, 9 H, tert-butyl), 0.66-1.97 (m, 21 H , cycloalkyl and hydroxyl), 3.13 (br m, 1 H, HCOH); mass spectrum (15 eV) 306 (21, M - H 2 0 ) , 234 (100, M - Me,SiOH). A solution of the crude 22 in 3.0 mL of T H F was treated with K H by the usual procedure. After 1.5 h at room temperature all the starting material was consumed. After the usual workup column chromatography (silica gel, hexanes) afforded 0.12 g (80% overall) of 23 as an oil and 0.02 g (10%) of starting silane (21), 23: IR (neat) 2950, 2850, 1485, 1395, 1250 cm-'; 'H N M R (CCI4) 6 0.87 (s, 9 H, tert-butyl), 0.57-2.80 (m, 20 H, cycloalkyl), 4.75-4.97 (d broadened by allylic coupling, J = 9 Hz, 1 H, vinyl); high resolution mass spectrum calcd for CI7H3,,: 234.2348. Found: 234.2345. 4-tert-Butyl-1-cyclopropylidenecyclohexane(29). A solution of 28 (0.1 1 g, 0.41 mmol) in 2 mL of diglyme was treated with KH by the standard procedure except that a reflux condenser was placed on the flask and the mixture was warmed to 90 "C for 3.5 h at which time the starting material was consumed. When diglyme is used as solvent, the following workup procedure is used. The cooled reaction mixture was partioned between hexanes and saturated NH,CI of pH 8. The organic layer was washed several times with saturated NH4Cl of pH 8 and dried (Na2S04), and the solvent was removed to yield a residue which, after purification by column chromatography (Silicar-CC-7, hexanes), afforded 0.062 g (86%) of 29 as an oil: IR (neat) 3000, 1400, 940 cm-I; 'H N M R (300 MHz, CDCI, + Me4Si) 6 0.85 (s, 9 H, tert-butyl), 0.95-1.25 (m, 7 H), 1.85 (m, 2 H), 2.00 (m, 2 H), 2.55 (m, 2 H); high resolution mass spectrum calcd for CI3Hz2178.1722, found 178.1722. 1-Cyclohexyl-2-methylpropene(17). A solution of 16 (0.20 g, 1.0 mmol) in 2 mL of diglyme was treated with K H by the standard procedure, and the reaction mixture was warmed to 90 "C. After 2 h, the starting material was consumed. The reaction mixture was worked up as in the previous diglyme reaction. Purification by column chromatography (Silicar-CC-7, hexanes) afforded 0.13 g (92%) of 17 whose spectral properties were identical with those of an authentic sample:53 'H N M R (CDCI,) 6 0.5-1.9 (m, 17 H, alicyclic H and allylic methyl groups), 4.8 (d broadened by allylic coupling, J = 9 Hz, 1 H, vinyl). 7-Hexylidenebicyclo[4.l.O]heptane (32). A solution of 31 (0.10 g, 0.40 mmol) in 5 mL of diglyme was treated with K H by the standard procedure. After 10 min the starting material was consumed. The reaction mixture was worked up as in the previous diglyme reaction. Purification by column chromatography afforded 0.064 g (98%) of 32 as an oil: IR
+
(52) Salaiin, J.; Hanack, M. J . Org. Chem. 1975, 40, 1774. (53) Farkas, J.; Novak. J. J. K. Collecf. Czech. Chem. Commun. 1960, 25, 1815.
3251
(neat) 3000, 1500, 1480, 820 cm-'; ' H N M R (CDCI,) 6 0.8-2.3 (m, 21 H, alicyclic and alkyl H), 5.8 (7, 1 H, vinyl); high resolution mass spectrum calcd for C13H22178.1722, found 178.1722. 7-(2-Methyl-2-propenylidene)bicyclo[4.1.O]heptane (34). A solution of 33 (0.10, 4.2 mmol) in 15 mL of T H F was treated with KH by the standard procedure. After 1.5 h the starting material was consumed. The solvent was removed from the washed and dried (Na2S04)organic layer yielding 0.64 g (100%) of 34 which was spectroscopically pure: IR (neat) 2950, 2850, 1620, 1450 cm-'; ' H N M R (CDC1,) F 1.0-2.0 (m, 13 H, alicyclic H and methyl), 5.0 (m, 2 H , gem-vinyl), 6.6 (m, 1 H, vinyl); high resolution mass spectrum calcd for C l l H 1 6148.1252, found 148.1252. 6-(2-Methyl-2-propenylidene)bicycl~3.I.O]hexane(37). A solution of 36 (1.14 g, 5.1 mmol) in 15 mL of T H F was treated with KH by the standard procedure. After 2 h the starting material was consumed. Purification by column chromatography (wash column using SilicarCC-7, hexanes) afforded 0.65 g (95%) of 37 as an oil: IR (neat) 3150, 3000,2900, 1640, 1480,900 cm-I; 'H N M R (CDCI,) 6 0.8-2.0 (m, 1 1 H, alicyclic H and methyl), 4.8 (m, 2 H , gem-vinyl), 6.4 (m, 1 H, vinyl); high resolution mass spectrum calcd for CloH14134.1096, found 134.1096.
cis- and tr~rans-@-Methoxyphenyl)[2-(2-methyl-l-propenyl)cyclopropylidene]methane (40). A solution of isomer A of 39 (1 15 mg, 0.378 mmol) in 1 mL of T H F was slowly added to a slurry of pentane-washed KH in 2 mL of T H F at 0 "C. The resulting mixture was allowed to warm to ambient temperature. After the solution had been stirred for 20 h, no starting material remained (TLC). The reaction mixture was worked up in the usual manner, and column chromatography (Silicar CC-7, 5% ethyl acetate-hexanes, R,0.55) afforded 76 mg (94%) of 40 as a pale yellow oil. This olefin decomposes rapidly in the presence of CDC1,; thus the N M R spectrum must be taken in CC1,: IR (neat) 2950, 2900, 1610, 1510, 1250, 1170, 1040, 860, 840 cm-I; ' H N M R (CCl,) 6 1.0-1.4 (m, 2 H, cyclopropyl), 1.7 (s, 3 H, methyl), 1.9 (s, 3 H , methyl), 2.1-2.3 (m, 1 H , tertiary cyclopropyl), 3.7 (s, 3 H , OCH3), 4.5-4.7 (m, 1 H, vinyl), 6.3-6.5 (m, 3 H, aromatic and benzylic vinyl), 7.1-7.3 (d, J = 8 Hz, 2 H, aromatic); high resolution mass spectrum calcd for ClsHlsO214.1358, found 214.1356. A solution of a mixture of isomers B, C, and D of 39 (1 15 mg, 0.378 mmol) in 1 mL of T H F was treated, worked up, and chromatographed as for isomer A to afford 77 mg (95%) of a pale yellow oil with the same spectral properties as that of the olefin derived from A. It was not determined whether 40 is a mixture of cis and trans olefins or a single compound. frans-l-(l-Cyclohexenyl)-2-cyclohexylether1e (43). In a 2-necked flask equipped with an argon inlet, a solution of 42 (0.47 g, 1.7 mmol) in 10 mL of CClp was stirred and cooled to 0 "C. Freshly distilled thionyl chloride (0.16 mL, 0.26 g, 2.1 mmol) was slowly added, and the resulting solution was stirred for 2 h at 0 "C. The CCI, and excess thionyl chloride were removed in vacuo, and the crude chloride was taken up in 10 mL of 1:l Me2SO/THF and rapidly stirred. Tetra-n-butylammonium fluoride (1.85 mL of a 1.00 M solution in THF, 1.85 mmol) was added to the crude chloride solution, and the mixture was stirred for 1.5 h and partitioned between water and hexanes. Column chromatography (Silicar-CC-7, hexanes) of the dried (Na,SO,) organic layer afforded 0.23 g (72%) of 43 and 0.07 g (15%) of cis- and trans-46. 43: IR (neat) 2900, 2850, 1450,950 cm-'; ' H N M R (300 MHz, CDCI,) 6 1.08-2.12 (m, 19 H, alicyclic), 5.51 (d of d, J = 15.8 H z and 7.0 Hz, 1 H , 2-ethenyl H ) , 5.65 (m, 1 H , cyclohexenyl), 6.01 (d, J = 15.8 Hz, 1 H, 1-ethenyl H); high resolution mass spectrum calcd for C,,H,, 190.1721, found 190.1721. 46: IR (neat) 2910, 2840, 1430, 1230, 840, 820 cm-'; ' H N M R (CDCI,) 6 0.00-0.10 (s, 9 H , epimeric CH,Si), 0.80-2.10 (m, 19 H), 5.05-5.15 (m. 1 H , endocyclic = CH), 5.40-5.65 (d, J = 1 1 Hz, 1 H , ethenyl H); mass spectrum (15 eV) 262 (100, M), 247 ( I O , M CH,), 188 (64, M - CH,SiH).
Acknowledgment. W e t h a n k Dr. Alvin M a r c u s for providing t h e mass spectral d a t a , t h e N a t i o n a l Institutes of H e a l t h for providing financial support (GM 22760), and the National Science Foundation for funds used t o purchase t h e 300-MHz Bruker NMR instrument used in this study (CHE 7905185). Registry No. 1 (R'R' = (CHJ5), 37457-08-2; 1 (R1R2= (CH,),), 69519-84-2; 7,71987-46-7; 8,69190-57-4; 10,76200-34-5; 11, 8965693-9; 12,51 197-95-6; 13,89656-94-0; 14,89656-95-1; 15,89656-96-2; 16, 89656-97-3; 17, 89656-98-4; 18, 89656-99-5; 19, 89657-00-1; 20, 27428-33-7; 21, 89657-01-2; 22, 89657-02-3; 23, 89657-03-4; 24, 74379-74-1; 25, 89657-04-5; 26, 55088-84-1; 27, 89657-05-6; 28, 89657-06-7; 29,89657-07-8; 30 (dithioacetal), 8371 1-04-0; 30 (7-endoMe,%), 89657-08-9; 30 (7-exo-Me3Si),89657-09-0; 31,8 1236-81-9; 32, 81236-88-6; 33,89675-83-2; 34,89657-10-3; 35 (dithioacetal), 5868 I -
3252
J. Am. Chem. SOC.1984, 106, 3252-3257
16-6; 35 (6-endo-Me3Si),89657-1 1-4; 35 (6-exo-Me3Si), 89657-12-5; 36, 89657-13-6; 37, 89657-14-7; cis-38, 89657-15-8; trans-38, 89657-16-9; cis-39 (isomer l ) , 89657-17-0; cis-39 (isomer 2), 89708-66-7; trans-39 (isomer l ) , 89708-67-8; trans-39 (isomer 2), 89708-68-9;40, 89657-18-1; 41, 71342-13-7; 42, 89657-19-2; 43, 53723-46-9; (E)-46, 89657-21-6; (2)-46, 89657-22-7; 48, 87729-87- 1; LDMAN, 74379-76-3; (CH3),C(SPh)2, 14252-46-1; (CH3)2C(OMe)2,77-76-9; CH3CH,CH(SPh)2,
15486-58-5; CH3CH2CH0, 123-38-6; c-C6H,,CH(SPh)2, 54905-14-5; c-C6H,,CH0, 2043-61-0; Me3SiC1, 75-77-4; C H 3 C H 0 , 75-07-0; MeOC6H,-p-CH0, 123-11-5; n-C6H&HO, 11 1-71-7; n-C5H,,CH0, 66-25-1; CH2=C(CH3)CH0, 78-85-3; KH, 7693-26-7; 1-(dimethylamino)naphthalene, 86-56-6; l,l-bis(phenylthio)-4-tert-butylcyclohexane, 85895-63-2; 4-tert-butylcyclohexanone, 98-53-3; exo-7-(phenylthio)norcarane, 89657-20-5; endo-7-(phenylthio)norcarane, 37942-20-4.
Stereocontrolled Total Synthesis of (-)- Maytansinol Masato Kitamura, Minoru Isobe,* Yoshiyasu Ichikawa, and Toshio Goto Contribution from the Laboratory of Organic Chemistry, Faculty of Agriculture, Nagoya University, Chikusa, Nagoya 464, Japan. Received September 14, 1983. Revised Manuscript Received January 25, 1984
Abstract: Chiral maytansinol (1) was synthesized stereoselectively from D-mannose. Asymmetric centers of 1 were induced intramolecularly from an asymmetric carbon corresponding to C-7 via the common key intermediate 3 for maytansinoids. Key points are (i) the usage of a carbohydrate as a chiral template for acyclic asymmetric induction and (ii) a remote stereochemical control in aldol reaction to elaborate the C-3 asymmetric center by intramolecular asymmetric induction.
Maytansinol (1) is the key compound in preparing ansa macrolactams such as maytansine (2a) found in Muytenus serrata,'
Scheme I" HOCH2
TrOCH,
RlOCHr
e 4
1
6
5
R=H
2 a R = COCHMeNMeAc b R=COCHMe2
A D Me I
I
I
OR?
I
I
PhVO
OR
9
6
/ 11
3
ansamitosin (2b) found in Nocardia sp. C - l 5 0 0 3 ( N - l ) , * and others, which have remarkable antitumor activity with different accompanying toxicity depending upon the ester side chain at C-3 p o s i t i ~ n . ~We have recently reported the total synthesis of racemic maytansinol in a highly stereocontrolled manner involving diastereotopic induction of all the asymmetric centers of (&)-l from only one asymmetric carbon corresponding to C-7.4 This success prompted us to the studies on t h e synthesis of optically active 1
a R , = T l i P , R , = h5e b R , = t i . R,= -(CH,)CI
R ' Cli3 R'
0 R'
CH:OAr
CHlO'.'cl.
e
k
-
3 /
Q
@PPh,
(a) (CH,),C(OMe),-PPTS, TsCl/Py, NaH-MeI; (b) H+, TrC1-Py ; (c) t-BuOK/THF; (d) super hydride, BnBr-NaH, H+;(e) (COCl),/ Me,SO/Et,N; (f) PhS(Me,Si),CLi/THF, rn-CPBA; (g) MeLi/THF. KF-MeOH, H ,/Pd-black, DHP-PPTS ; (h) n-Bu Li-4-bromopent-2ene/THF-HMPA, CSA/HOCH,CH,Cl; (i) CrO,-ZPy, CSA/HC(OMe),-MeOH, PhSNa, m-CPBA; (j) NaBH,, AcCl/Py, t-BuMe,SiC1-imidazole/DMF, O,/CH,Cl,, Et,N; ( k ) PPTS/CH(OMe),MeOH, NaOMe/MeOH, CrO,-2Py, lS/THF-DMF ( 2 : 1 ) , AcOHTHF-H20 ( 4 : l : I ) .
1979, 35, 1079. (3) (a) Komoda, Y . ; Kishi, T. 'Anticancer Agents Based on Natural Product Models"; Ed. Cassady, J. M., Douros, J. D., Eds.; Academic Press: New York, 1980; Chapter 10, pp 353-389. (b) Corey, E. J.; Hua, D. H.; Seitz, S. P. Tetrahedron Letr. 1984, 25, 3. (4) (a) Isobe, M.; Kitamura, M.; Goto, T. J . Am. Chem. SOC.1982, 104, 4997. (b) Kitamura, M.; Isobe, M.; Ichikawa, Y.; Goto, T., manuscript submitted for publication in J . Org. Chem. (c) Other synthetic works, see the following papers and the references cited therein: Meyers, A. I.; Reider, P. J.; Campbell, A. L. J . Am. Chem. SOC.1980, 102,6597. Corey, E. J.; Weigel, L. 0.;Chamberlin, A. R.; Cho, H.; Hua, D. H. J . Am. Chem. SOC.1980,102, 6615. I
E
R ? CHO
(1) Kupchan, S. M.; Komoda, Y.; Branfman, A. R.; Sneden, A. T.; Court, W. A,; Thomas, G. J.; Hintz, H. P. J.; Smith, R. M.; Karim, A,; Howie, G. A.; Verma, A. K.; Nagao, Y.; Dailey, R. G., Jr.; Zimmerly, V. A.; Sumner, W. C. Jr. J . Org. Chem. 1977, 42, 2349. (2) Asai, M.; Mizuta, E.; Izawa, M.; Haibara, K.; Kishi, T. Tetrahedron
0002-7863/84/1506-3252$01.50/0 , ,
a b R = -,CH:)SPh - i c i i ICI C R = -ICtI,ISO,Ph
via t h e s a m e key intermediate 3 a s t h e racemic case. Synthesis of the Chiral Key Intermediate (-)-3. T h e key intermediate (-)-3 was prepared from D-mannose (4a) both as the chiral starting material and as the chiral template for the induction
0 1984 American Chemical Societv -
